Inducible systems that perturb the activity of cell signaling molecules are powerful tools for probing pathway dynamics and dependencies in living cells and animals. Photoactivation, or caging, is an excellent method for inducing changes because it can be nearly instantaneous and activation can be spatially localized. Photoactivation of proteins has generally required site-specific chemical modification that is performed in vitro, generating analogs that are often difficult to add to cells and are irreversibly activated. Our goal is to create photoactivatable proteins that are genetically encodable, and therefore, can be readily introduced into living cells by DNA transfection. Our design strategy makes use of the naturally photoreactive LOV2 domain from the plant protein phototropin. When activated with blue light, the flavin chromophore in the LOV2 domain forms a covalent bond with cysteine 450, creating a structural perturbation that leads to the unfolding of the C- terminal helix of the LOV2 domain (the J1-helix). We will test if the light mediated unfolding of the LOV2 J1- helix can be used to control the activities of proteins or peptides that are either fused to or embedded within the J1-helix. We will focus on caging proteins and peptides that activate critical signaling pathways in cell migration. In aim 1, fusions with the LOV2 domain will be used to create photoactivatable variants of the small GTPases Rac1, Cdc42 and RhoA. Preliminary studies indicate that caging requires favorable interactions between surface residues on the GTPase and the LOV2 domain. A crystal structure of a LOV2-Rac1 fusion will be used as a template for protein design simulations to identify mutations that stabilize the caged state of LOV2-GTPase fusions. In aim 2, multi-state protein design simulations will be used to vary the sequences of naturally occurring peptide activators and inhibitors so that they can be embedded in the folded J1-helix in the dark state, but still bind their target proteins in the lit state. In aim 3, we will test if photoactivable LOV2 variants and their binding partners can be used as modules for inducing the dimerization of signaling molecules. These studies will reveal general strategies for the photoactivation of proteins with the LOV2 domain as well as provide powerful tools for studying a variety of cellular processes.